A New Integrating Multi-features C5.0 Decision Tree Method for Classification of Tropical Rubber Woods Using High-resolution Remote Sensing Image
Abstract
In order to classify rubber woods using high-resolution remote sensing images, an integrating multi-features C5.0 decision tree (abbreviated as IMFC5) classification method is proposed. The procedure of IMFC5 is as follows: (1) images are pre-processed, all features are collected and integrated into one file; (2) decision features are obtained by the use of BFDFS method, and the classification rules are built; (3) the output is obtained after executing the classification rules and carrying out classification post-processing. The proposed method is tested in an experimental area of Guangba Farm, Dongfang City, Hainan Island, China. The results indicate that producer accuracy, user accuracy, total accuracy of rubber woods, and Kappa coefficient are 88%, 91.67%, 92%, and 0.89, respectively. All four indices are better than the other classification methods’, proving feasibility and efficiency of the proposed method
Keywords
Classification of rubber woods, High-resolution remote sensing image, Integrating multi-features C5.0 decision tree (IMFC5) method.
DOI
10.12783/dtcse/smce2017/12425
10.12783/dtcse/smce2017/12425
Refbacks
- There are currently no refbacks.